Measurement of tissue metabolite distributions by 1H MR Spectroscopic Imaging (MRSI) provides a sensitive diagnostic neuroimaging modality; however, it remains underutilized in the clinical setting. One requirement that has limited a wider use of these techniques is the need for comprehensive data processing procedures, which optimally make extensive use of prior information, including tissue and morphological analysis from MRI and knowledge of normal metabolite distributions. In the previous funding period this project has developed an integrated software package that provides comprehensive and automated MRI and MRSI processing that includes signal normalization and spatial transformation, thereby enabling formation of a reference image database of MR-observed human metabolite values as a function of acquisition, spatial, and subject parameters. This database enables mapping of brain metabolite distributions and provides valuable normative information for analysis of individual patient studies. In this project period these processing methods will be further developed and integrated with standardized MRSI acquisition protocols implemented on MR instruments from three major manufacturers, and the metabolite database will be expanded to support the developed acquisition protocols. This project will therefore develop standardized 1H MRSI protocols, including cross-site and long-term validation and quality control, that will maximize the impact of MRSI for clinical studies and facilitate multi-site quantitative metabolic imaging studies. The effectiveness of these standardized metabolic neuroimaging protocols will be evaluated by installing a turnkey system at a clinical research site where performance metrics will be developed and data acquired to test the sensitivity of the standardized MRSI methods for detection of metabolic changes in individual subjects. All developments will be made freely available to researchers involved in the application of MRSI for clinical diagnostic imaging. The bioengineering focus areas of this project include signal and image processing, MR neuroimaging sequence development, multiparametric image analysis methods, and networked image databases. The partnership sites are the University of Miami (leading institution), University of California Los Angeles, Johns Hopkins University, Stanford University, and Duke University. This development will improve the quality of the diagnostic neuroimaging information obtained using MR spectroscopy; provide cost savings by sharing data from normal subject; establish standardized metabolic imaging methods; and map metabolite distributions in human brain.